Improvement of Deconvolution Algorithm for Space Mission
Budget: €20,000 – €50,000 EUR
Request for Offer (RFO) for Software Engineer Services for and Space Mission Project
Project Overview
Details of the software
The developed Python software (as part of this service) will be used to process astronomical data by decreasing the diffusion of light within the images produced by an space mission telescope. This diffusion of light has the effect of smoothing the images. The main steps of the required ervice include:
1. Algorithm development: some basic deconvolution algorithms have been implemented within the core tools used within the scientific project pipeline. But the following features are not yet implemented in a way that can be useful within the mission pipeline.
- They are not optimized to the low surface brighntess (lower signal-to-noise ratio) regions.
- The algorithms currently do not match an image that has a wide point spread function (PSF), to a sharper pre-defined PSF.
- They are applied to a single (already coadded/deep image) image. But deconvolution methods have recently been proposed to be applied to single exposures that can optimize the coadding process within the pipeline.
2. Testing and Validation: each developed algorithm from the previous step will be tested on the following images:
- Simulated data. The simulated exposures have become very realistic, and provide us with an ideal data set where we know the ground-truth to evaluate the success of the algorithms.
- Real data: We will try many aspects of the algorithms developed in step one on public data.
3. Optimizations and execution
- Testing optimizations with GPUs
- Execution within in the complete the reduction pipeline on the high performance computing (HPC) environment, access provided by the client.
Requirements and Deliverables:
The software will be required for September 2025. The provider will need to integrate the code in our open source library tool. For that it will require regular meetings with key members of the collaboration. The service should provide 3 deliverables:
1. Initial Phase – Requirements & High-Level Design (Milestone 1)
• Deliverable: Initial flow diagram showing the general architecture and main processes.
• Milestone: Completion of requirements gathering and system architecture design.
• Goal: Ensure mutual understanding of functionalities, user interactions, and system components.
2. Midway Phase – Development & Integration (Milestone 2)
• Deliverable: Updated flow diagram reflecting refinements based on progress and feedback.
• Milestone: Completion of a functional prototype or beta version.
• Goal: Validate core functionalities, integration feasibility, and potential adjustments.
3. Final Phase – Testing & Deployment (Milestone 3)
• Deliverable: Final flow diagram, fully aligned with the implemented system.
• Milestone: Completion of final software version and validation.
• Goal: Deliver tested, stable, and documented software ready for deployment.
For those interested we will need a budget for the three deliverables by the 15th of March. Please reach out if you need more details.
Project Overview
Details of the software
The developed Python software (as part of this service) will be used to process astronomical data by decreasing the diffusion of light within the images produced by an space mission telescope. This diffusion of light has the effect of smoothing the images. The main steps of the required ervice include:
1. Algorithm development: some basic deconvolution algorithms have been implemented within the core tools used within the scientific project pipeline. But the following features are not yet implemented in a way that can be useful within the mission pipeline.
- They are not optimized to the low surface brighntess (lower signal-to-noise ratio) regions.
- The algorithms currently do not match an image that has a wide point spread function (PSF), to a sharper pre-defined PSF.
- They are applied to a single (already coadded/deep image) image. But deconvolution methods have recently been proposed to be applied to single exposures that can optimize the coadding process within the pipeline.
2. Testing and Validation: each developed algorithm from the previous step will be tested on the following images:
- Simulated data. The simulated exposures have become very realistic, and provide us with an ideal data set where we know the ground-truth to evaluate the success of the algorithms.
- Real data: We will try many aspects of the algorithms developed in step one on public data.
3. Optimizations and execution
- Testing optimizations with GPUs
- Execution within in the complete the reduction pipeline on the high performance computing (HPC) environment, access provided by the client.
Requirements and Deliverables:
The software will be required for September 2025. The provider will need to integrate the code in our open source library tool. For that it will require regular meetings with key members of the collaboration. The service should provide 3 deliverables:
1. Initial Phase – Requirements & High-Level Design (Milestone 1)
• Deliverable: Initial flow diagram showing the general architecture and main processes.
• Milestone: Completion of requirements gathering and system architecture design.
• Goal: Ensure mutual understanding of functionalities, user interactions, and system components.
2. Midway Phase – Development & Integration (Milestone 2)
• Deliverable: Updated flow diagram reflecting refinements based on progress and feedback.
• Milestone: Completion of a functional prototype or beta version.
• Goal: Validate core functionalities, integration feasibility, and potential adjustments.
3. Final Phase – Testing & Deployment (Milestone 3)
• Deliverable: Final flow diagram, fully aligned with the implemented system.
• Milestone: Completion of final software version and validation.
• Goal: Deliver tested, stable, and documented software ready for deployment.
For those interested we will need a budget for the three deliverables by the 15th of March. Please reach out if you need more details.